Re: EWW: initially failed loading of images
Eli Zaretskii <[email protected]> Sat, 16 May 2026 17:46:11 +0300
| Newsgroups | gmane.emacs.help |
|---|---|
| Message-ID | <[email protected]> |
> Date: Sat, 16 May 2026 08:47:40 +0300 > From: Eli Zaretskii <[email protected]> > Cc: [email protected], [email protected], > [email protected] > > > From: [email protected] > > Cc: [email protected], [email protected], Christopher Howard > > <[email protected]> > > Date: Fri, 15 May 2026 12:46:12 -0400 > > > > I have been able to reproduce the problem using Emacs 30, in > > case Christopher Howard wants to report it (using M-x > > report-emacs-bug). > > > > 1. emacs -Q > > 2. (setopt eww-retrieve-url 'sync) > > 3. M-x eww RET > > 4. At the prompt "Enter URL or keywords: ", enter https://en.wikipedia.org/wiki/Submarine > > 5. Scroll down many screens (the initial Wikipedia logo is > > displayed) until the first submarine image is displayed. > > Below it is a gray box for a missing image. > > 6. Type ‘g’, which is bound to the command ‘eww-reload’, by > > default. > > 7. After the page is reloaded, the second image will be > > displayed (another submarine). > > 8. Scroll many more screens to see that some additional > > images have been downloaded, but many gray boxes still > > appear. It then requires multiple reloads to get all of > > the images to download. > > Please submit a full bug report with these details, and thanks. AFAICT, the problem is that we hit HTPP error 429 (too many requests). This explains why one needs to type 'g' many times, and each time we get about 4 more images, before we hit 429 again. Someone who knows about this stuff should adapt shr.el to the current DDoS hysteria of the servers out there, because shr.el just issues another url-fetch request each time it sees an image reference in the document.